Simple pricing for Moondream.

Models and Photon are free. Cloud and Lens are pay-per-token. Paid plans add collaboration and support.

Pick what you need.

Photon
Free

The world's fastest VLM inference engine — runs everywhere, from edge to server-class hardware.

$npm install moondream
Models
Free

Open and free to use — hosting them as a cloud service requires our license.

..or click here if you really want to download them: Moondream 2 · Moondream 3 (preview)

Cloud
Pay per token

Hosted Moondream inference — lightning fast, always on, with HIPAA and SOC 2 available.

Real-time inference
$0.30 / 1M input tokens
$2.50 / 1M output tokens
Batch inference
$0.15 / 1M input tokens
$1.25 / 1M output tokens
Lens
Pay per token

Fine-tune Moondream for your task.

Rollouts
Same as Cloud inference
Training
$0.60 / 1M training tokens

Need collaboration or support?

Paid plans add seats, shared resources, billing, and direct support. They do not change token pricing.

Free
$0/ month
 
Best for individual developers.
Start free
  • 1 seat
  • Personal workspace
  • Discord support
  • $5/month usage credits
Recommended
Team
$350/ month
or $300/mo billed annually
Best for teams building together.
Start Team
  • 10 seats
  • Shared fine-tunes
  • Shared projects
  • Shared API keys
  • Shared billing
  • Dedicated Slack
  • 1h/month consulting
  • $5/month usage credits
Additional seats available. Contact us.
Enterprise
Custom
 
Best for larger, regulated, or custom deployments.
  • Custom seats
  • HIPAA / BAA available
  • Security review
  • Priority support
  • Custom deployment support
  • Custom billing

Paid plans do not change Cloud or Lens token pricing.


Good to know

  • Every workspace includes $5/month usage credits.
  • Fine-tuned models use the same inference pricing as base models.
  • Fine-tunes are unlimited. Usage and rate limits still apply.
  • Paid plans are for collaboration and support, not cheaper tokens.